Abstract This administrative supplement will support the development of Rosaura Romero's research and professional skills and provide her with mentoring for graduate school. We are not proposing any changes to the scope or human subjects of the parent grant and the abstract remains as follows: Lower-income and minority groups face significant health disparities with respect to obesity, cancer, heart disease and other diet-related chronic conditions. Poor diets, low in fruits and vegetables (F&V) and high in saturated fat, sodium and sugar, contribute to many of the health problems faced by vulnerable groups. While socioeconomic status and other individual level factors (i.e., food preferences, time and skills to prepare healthy food, etc.) can lead to reduced F&V consumption, these must be viewed in an environmental context. Compared with higher-income neighborhoods, lower-income and minority neighborhoods are less likely to have stores that sell a variety of F&V and other healthy foods. When stores are available, produce may not be affordable, high quality or culturally appropriate. Farmers' markets and mobile produce markets (MM) have become increasingly popular strategies to alleviate food access concerns in underserved communities. However, it is unclear if these programs have the necessary components to have an appreciable impact on diet. Our research team recently completed one of the first randomized controlled trials of a MM program called the Veggie Van (VV). VV delivered boxes of fresh, locally grown produce and food-focused education to communities with significant barriers to F&V consumption including availability, affordability, quality and knowledge. In this small cluster- randomized trial in 12 communities (N=201), we saw impressive changes in F&V intake with intervention participants eating almost 1 more cups per day of F&Vs than the control group. Intervention participants also reported increases in perceived access to healthy foods and VV customers attributed many dietary changes to the MM program. While these results are very promising, we believe it is important to test the effectiveness of the VV program when implemented by different organizations in multiple communities. If shown to be effective, we can create a research-tested intervention toolkit which can be disseminated to communities across the country. For this research, we identified 8 organizations nationwide that are well-qualified to implement the VV model. Organizations will identify appropriate sites for MM deliveries (32 total) and we will randomize them to either an implementation or planning condition. With the help of our team's technical assistance and provided funding, partner organizations will engage community members in the process and initiate a MM program. We will use a Type 1 Hybrid Effectiveness-Implementation to measure effectiveness (diet, BMI, dermal carotenoids) and implementation (customer reach and sales, process measures, qualitative interviews with MM staff). We will also examine sustainability of MM financial models and determine implementation standards (i.e., dose needed to maintain impact) for inclusion in our MM toolkit for future dissemination.